The resolution of acyclic P-stereogenic phosphine oxides via the formation of diastereomeric complexes: A case study on ethyl-(2-methylphenyl)-phenylphosphine oxide

Article abstract of DOI:10.1002/chir.22816

As an example of acyclic P-chiral phosphine oxides, the resolution of ethyl-(2-methylphenyl)-phenylphosphine oxide was elaborated with TADDOL derivatives, or with calcium salts of the tartaric acid derivatives. Besides the study on the resolving agents, several purification methods were developed in order to prepare enantiopure ethyl-(2-methylphenyl)-phenylphosphine oxide. It was found that the title phosphine oxide is a racemic crystal-forming compound, and the recrystallization of the enantiomeric mixtures could be used for the preparation of pure enantiomers. According to our best method, the (R)-ethyl-(2-methylphenyl)-phenylphosphine oxide could be obtained with an enantiomeric excess of 99% and in a yield of 47%. Complete racemization of the enantiomerically enriched phosphine oxide could be accomplished via the formation of a chlorophosphonium salt. Characterization of the crystal structures of the enantiopure phosphine oxide was complemented with that of the diastereomeric intermediate. X-ray analysis revealed the main nonbonding interactions responsible for enantiomeric recognition.

Full text of DOI:10.1002/chir.22816

Received: 3 November 2017
DOI: 10.1002/chir.22816
Revised: 6 December 2017
Accepted: 8 December 2017
R E G U L A R A R T I C L E
The resolution of acyclic P‐stereogenic phosphine oxides via
the formation of diastereomeric complexes: A case study on
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide
Péter Bagi1
| Bence Varga1 | András Szilágyi2 | Konstantin Karaghiosoff3 | Mátyás Czugler1 |
Elemér Fogassy1 | György Keglevich1
1 Department of Organic Chemistry and
Abstract
Technology, Budapest University of
Technology and Economics, Budapest,
Hungary
As an example of acyclic P‐chiral phosphine oxides, the resolution of ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide was elaborated with TADDOL deriva-
tives, or with calcium salts of the tartaric acid derivatives. Besides the study
on the resolving agents, several purification methods were developed in order
to prepare enantiopure ethyl‐(2‐methylphenyl)‐phenylphosphine oxide. It was
found that the title phosphine oxide is a racemic crystal‐forming compound,
and the recrystallization of the enantiomeric mixtures could be used for the
preparation of pure enantiomers. According to our best method, the (R)‐
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide could be obtained with an
enantiomeric excess of 99% and in a yield of 47%. Complete racemization of
the enantiomerically enriched phosphine oxide could be accomplished via the
formation of a chlorophosphonium salt. Characterization of the crystal struc-
tures of the enantiopure phosphine oxide was complemented with that of the
diastereomeric intermediate. X‐ray analysis revealed the main nonbonding
interactions responsible for enantiomeric recognition.
2 Department of Physical Chemistry and
Material Science, Budapest University of
Technology and Economics, Budapest,
Hungary
3 Department Chemie und Biochemie,
Ludwig‐Maximilians‐Universität
München, Munich, Germany
Correspondence
Péter Bagi, Department of Organic
Chemistry and Technology, Budapest
University of Technology and Economics,
Műegyetem rkp. 3., H‐1111 Budapest,
Hungary.
Email: pbagi@mail.bme.hu
Funding information
New National Excellence Program of the
Ministry of Human Capacities, Grant/
Award Number: ÚNKP‐17‐4‐I‐BME‐102;
National Research, Development and
Innovation Office—NKFIH, Grant/Award
Number: OTKA PD 116096
KEYWORDS
acyclic phosphine oxide, enantiomeric enrichment, optical resolution, P‐stereogenic center,
X‐ray structure
1 | INTRODUCTION
protocols were developed for the preparation of the
corresponding P(III) compounds, and the stereochemi-
Since their first appearance, P‐stereogenic ligands form an
important class of compounds used in asymmetric cataly-
sis,1,2 and there is still a continuing interest for novel syn-
thetic methods, which can be applied for the preparation
of optically active P‐chiral derivatives.
The synthesis of optically active P‐chiral compounds
generally involves the preparation of the corresponding
phosphine boranes or phosphine oxides, as they are
regarded as bench stable precursors of phosphines.3,4
Moreover, a number of deoxygenation or deboronation
cal outcome of those reactions was also studied in
detail.5,6
Enantioselective synthesis is one of the main sources
for optically active organophosphorus compounds.7-9 In
those reactions, (−)‐menthol and (−)‐ephedrine are fre-
quently used chiral templates,6,10-17 but other synthetic
derivatives were also used.18-21 Enantioselective synthesis
affords enantiomerically enriched or enantiopure com-
pounds. However, in this case, the purification of the
enantiomeric mixtures should be an integral part of
Chirality. 2018;30:509–522.
wileyonlinelibrary.com/journal/chir
© 2018 Wiley Periodicals, Inc.
509

510
BAGI ET AL.
these procedures. The self‐disproportion of enantiomeric
mixtures via achiral chromatography or sublimation is
receiving more and more attention,22-26 but the first
example of this phenomenon in the sphere of P‐stereo-
genic compounds is yet to be published.
2 | MATERIALS AND METHODS
2.1 | General (instruments)
The 31P, 13C, and 1H NMR spectra were taken on a Bruker
AV‐300 or DRX‐500 spectrometer operating at 121.5, 75.5,
and 300 or 202.4, 125.7, and 500 MHz, respectively. Cou-
pling constants are expressed in hertz.
The exact mass measurements were performed
using a Q‐TOF Premier mass spectrometer in positive
electrospray mode.
The enantiomeric excess (ee) values were determined
by chiral high‐performance liquid chromatography
(HPLC) on a PerkinElmer Series 200 instrument equipped
with chiral HPLC using Kromasil 5‐Amycoat column
(250 × 4.6‐mm ID, hexane/ethanol 85/15 as an eluent
with a flow rate of 0.8 mL/min, T = 20°C, UV detector
α = 254 nm). Retention times: 8.4 minutes for (S)‐3 and
12.6 minutes for (R)‐3.
The other main approach for the preparation of opti-
cally active P‐chiral compounds involves the separation
of racemic mixtures. One elegant possibility is preparative
chromatography on a chiral stationary phase.27 This
approach was used for the preparation of a few
enantiopure P‐chiral phosphine oxides on a multigram
scale.28-33 Moreover, the costs of such separations may
be decreased by using a supercritical fluid as the mobile
phase, which complements the time efficiency of chro-
matographic separations with cost efficiency.34 Optical
resolution via the formation of diastereomers is the most
common way for the separation of racemic mixtures into
enantiomers. Many methods based on the formation of
covalent diastereomers, diastereomeric salts, or diastereo-
meric complexes were elaborated for P‐chiral phosphine
oxides.3,4,35-39 In a few cases, dynamic resolutions were
also developed. These methods involved in situ racemiza-
tion of the phosphine oxide or its derivative under the res-
olution conditions. In this manner, the total amount of
the racemic starting material may be converted into a sin-
gle enantiomer.40-44 However, many enantioseparation
methods developed for chiral phosphine oxides comprise
classical resolutions. Although a number of methods were
elaborated over the decades, most of these resolutions
were individual examples, and a generally applicable
resolving agent for a wide variety of phosphine oxides is
still to be found.3,4 O,O′‐dibenzoyl‐tartaric acid was one
of the most commonly applied resolving agents, but the
main scope of substrates comprised a few P‐chiral
bis(phosphine oxides),30,45-47 and species bearing an axi-
ally dissymmetric element.48-53 Moreover, this resolving
agent could also be used for the enantioseparation of
tert‐butyl‐phenylphosphine oxide.44,54 Our research group
found that TADDOL derivatives and the Ca2+ salts of
tartaric acid derivatives were suitable resolving agents to
separate the enantiomers of several 5‐ and 6‐membered
P‐heterocyclic phosphine oxides having various substitu-
tion patterns.55-65
Optical rotations were determined on a PerkinElmer
241 polarimeter.
Differential scanning calorimetry (DSC) curves were
recorded by using a PYRIS Diamond Differential Scanning
Calorimeter (PerkinElmer). The measurements were per-
formedon~10‐mgsamplesinN2 atmospherewithaheating
rate of 10°C/min between 30°C and 200°C. The tempera-
tures at the minimum of the endotherms were referred to
as the melting points of the enantiomeric mixtures.
Melting points were determined in capillary tubes by
using a Gallenkamp melting point apparatus.
The microwave reaction was conducted in a 300 W
CEM Discover focused microwave reactor equipped with
a pressure controller using 20 to 30 W irradiation under
isothermal conditions.
The (−)‐(4R,5R)‐4,5‐bis(diphenylhydroxymethyl)‐2,2‐
dimethyldioxolane [(−)‐4], (−)‐(2R,3R)‐α,α,α′,α′‐tetraphenyl‐
1,4‐dioxaspiro[4.5]decan‐2,3‐dimethanol [(−)‐5] were
synthesized as described earlier.66 The (−)‐O,O′‐dibenzoyl‐
and (−)‐O,O′‐di‐p‐toluoyl‐(2R,3R)‐tartaric acid were pur-
chased from Sigma Aldrich.
2.2 | The preparation of ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide (3)
The aim of this study is to demonstrate that our
resolution methods can be applied to a wider substrate
scope. Herein, we report the resolution of ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide and the purifica-
tion strategies for the corresponding enantiomeric
mixtures, as a detailed example for the preparation of
optically active acyclic diaryl‐alkylphosphine oxides.
Moreover, the structures of the optically active phosphine
oxide prepared and one of its diastereomeric intermediate
were also elucidated by X‐ray crystallography.
To the solution of 9.0 g (46 mmol) phenylphosphonic
dichloride (1) in 75 mL of diethyl ether was added
dropwise 51 mmol of 2‐methylphenylmagnesium bromide
in 35 mL of diethyl ether [prepared from 1.3 g (53 mmol) of
magnesium and 6.1 mL (51 mmol) of 2‐bromotoluene]
over 30 minutes at −78°C. After the addition, the reaction
mixture was allowed to warm slowly to 26°C, and the mix-
ture was stirred further for 1 hour. Then, the reaction
mixture was cooled to −78°C, and 58 mmol of

BAGI ET AL.
511
TABLE 1 Resolution of ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) with TADDOL derivatives (−)‐4 and (−)‐5
Resolving
Agent
Diastereomer
complexb
Yieldc,f
(%)
eed,f
(%)
Abs.
Config.g
Entry
Eq.
Solventsa
Se,f (−)
1
TADDOL
0.5
2 × EtOAc/10 × hexane
(3)·(TADDOL)
(103) 74
(86) 35
(4) 3
(0.04) 0.02
(S)
2
3
4
spiro‐TADDOL
spiro‐TADDOL
spiro‐TADDOL
0.5
0.5
1
2 × EtOAc/10 × hexane
4 × iPrOH
(3)·(spiro‐TADDOL)
(3)·(spiro‐TADDOL)
(3)·(spiro‐TADDOL)
(6) 66
(19) 89
(11) 90
(0.05) 0.23
(0.22) 0.38
(0.17) 0.38
(R)
(R)
(R)
(117) 43
(148) 42
4 × iPrOH
Abbreviation: ee, enantiomeric excess.
aMixture of solvents for the crystallization and recrystallizations [milliliter of solvent/g of resolving agent].
bThe ratio of (3) and the resolving agent was determined by 1H NMR.
cThe yield of the diastereomer was calculated based on the half of the racemate (3) that is regarded to be 100% for each antipode.
dDetermined by chiral HPLC.
eResolving capability, also known as the Fogassy parameter [S = (Yield/100) × (ee/100)].67
fThe results obtained after the first crystallization are shown in parentheses, while the results obtained after 2 recrystallizations are shown in boldface.
gThe absolute configuration of (3) was determined by X‐ray analysis (vide infra).
ethylmagnesium bromide in 50 mL of diethyl ether [pre-
pared from 1.5 g (60 mmol) of magnesium and 4.3 mL
(58 mmol) of bromoethane] was added dropwise over
30 minutes. Then, the reaction mixture was allowed to
warm to 26°C, and the mixture was stirred overnight.
The reaction was then quenched with 80 mL of saturated
NH4Cl solution at 0°C. The 2 phases were separated, the
aqueous layer was extracted with diethyl ether. The
organic layers were washed with brine, and then dried
(Na2SO4). The crude product obtained after evaporating
the solvent was purified by column chromatography (silica
gel, 5% methanol in ethyl acetate) to give 5.5 g (49%) of
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) as a
white solid. Mp 87–88 °C; 31P NMR (121.5 Hz, CDCl3, δ):
36.2 (δlit 35.5)42; 1H NMR (500 MHz, CDCl3, δ): 7.67–7.60
(m, 3H, Ar‐H), 7.50–7.39 (m, 4H, Ar‐H), 7.30–7.26 (m,
1H, Ar‐H), 7.22–7.20 (m, 1H, Ar‐H), 2.43–2.19 (m, 5H,
CH2CH3 and Ar‐CH3), 1.19 (dt, J = 16.9 and 7.6, 3H,
CH2‐CH3); 13C NMR (125.7 Hz, CDCl3, δ): 142.7 (2JP‐C
= 7.6, C2’), 133.7 (1JP‐C = 96.5, C1), 132.1 (3JP‐C = 10.2,
C3’), 132.0 (4JP‐C = 2.6, C4’), 131.6 (2JP‐C = 11.1, C6’),
131.5 (4JP‐C = 3.1, C4), 130.9 (2JP‐C = 9.5, C2), 130.2 (1JP‐C
= not visible, C1’), 128.6 (3JP‐C = 11.6, C3), 125.5 (3JP‐C
= 11.8, C5’), 22.7 (1JP‐C = 73.4, CH2), 21.5 (2JP‐C = 4.3,
Ar‐CH3), 5.8 (2JP‐C = 4.9, CH2CH3); HRMS (ESI, m/z):
[M + H]+ calcd for C15H17OP 245.1090; found, 245.1085.
added. Colorless crystalline diastereomeric complex of
(R)‐3·(spiro‐TADDOL) appeared immediately. After
standing at 26°C for 3 hours, the crystals were separated
by filtration to give 0.20 g (86%) of (R)‐3·(spiro‐TADDOL)
with a diastereomeric excess (de) of 6%. The diastereo-
meric complex (R)‐3·(spiro‐TADDOL) was purified fur-
ther by 2 recrystallizations from a mixture of 0.31 mL of
ethyl acetate and 1.6 mL of hexane to afford 0.081 g
(35%) of the complex (R)‐3·(spiro‐TADDOL) with a de of
66% (Table 1, entry 2). The ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] was recovered from the
diastereomer by column chromatography (silica gel,
dichloromethane‐methanol 97:3) to give 0.023 g (31%) of
phosphine oxide (R)‐3 with an ee of 66%. The resolution
of racemic ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3) was also elaborated with spiro‐TADDOL [(−)‐5]
using isopropyl alcohol. According to this variation,
the mixture of the racemic compound (3) and spiro‐
TADDOL [(−)‐5] was dissolved in hot isopropyl al-
cohol, and the corresponding diastereomeric complex
(R)‐3·(spiro‐TADDOL) precipitated on cooling the mix-
ture to 26°C (Table 1, entries 3 and 4). Resolution of race-
mic ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3)
was also accomplished with TADDOL [(−)‐5] according
to the Representative Procedure I. (Table 1, entry 1). Mp
of (R)‐3·(spiro‐TADDOL) (de: 89%): 155°C to 157°C.
2.3 | Resolution of ethyl‐(2‐
2.4 | Resolution of ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide
(3) with spiro‐TADDOL [(−)‐5]
(Representative Procedure I)
methylphenyl)‐phenylphosphine oxide (3)
with calcium hydrogen O,O′‐dibenzoyl‐
(2R,3R)‐tartrate [(−)‐6] (Representative
Procedure II)
In 0.31 mL of hot ethyl acetate were dissolved 0.15 g
(0.61 mmol) of racemic ethyl‐(2‐methylphenyl)‐phenyl-
phosphine oxide (3) and 0.16 g (0.31 mmol) of spiro‐
TADDOL [(−)‐5], and then, 1.6 mL of hexane was
To 0.23 g (0.61 mmol) of DBTA·H2O in a mixture of
0.74 mL of ethanol and 0.075 mL of water was added
0.017 g (0.31 mmol) of CaO, and the mixture was heated

512
BAGI ET AL.
TABLE 2 Resolution of ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) with Ca(H‐DBTA)2 or Ca(H‐DPTTA)2 [(−)‐6 or (−)‐7]
Resolving
Entry Agent
Diastereomer
Complexb
Yieldc,f eed,f
(%) (%)
Abs.
Eq.
0.25 3 × EtOAc/3 × EtOH/ 10%H2O
Solventsa
Se,f (−)
Config.g
1
2
3
4
5
6
Ca(H─DBTA)2
Ca(H─DBTA)2
Ca(H─DBTA)2
Ca(3)2(H─DBTA)2
Ca(3)2(H─DBTA)2
Ca(3)(H─DBTA)2
Ca(3)2(H─DPTTA)2
(61) 25 (61) 94 (0.37) 0.23
(79) 14 (82) 87 (0.64) 0.12
(57) 15 (64) 94 (0.36) 0.14
(63) 18 (58) 82 (0.36) 0.15
(R)
(R)
(R)
(R)
(R)
(R)
0.5
0.5
3 × EtOAc/3 × EtOH/ 10%H2O
3 × MeCN/3 × EtOH/ 10%H2O
Ca(H─DPTTA)2 0.25 3 × EtOAc/3 × EtOH/ 10%H2O
Ca(H─DPTTA)2 0.5
Ca(H─DPTTA)2 0.5
3 × EtOAc/3 × EtOH/ 10%H2O Ca3(3)4(H─DPTTA)6 (59) 8
(69) 83 (0.41) 0.07
3 × MeCN/3 × EtOH/ 10%H2O Ca(3)2(H─DPTTA)2 (76) 14 (76) 88 (0.54) 0.13
Abbreviation: ee, enantiomeric excess.
See Table 1 for footnotes.
at the boiling point until it became clear. To the solution of
the in situ formed resolving agent Ca(H‐DBTA)2 [(−)‐6]
was then added 0.15 g (0.61 mmol) of racemic ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide (3) in 0.74 mL of
ethyl acetate. After the addition, the solution was allowed
to cool down to 26°C, whereupon colorless crystals appeared.
After standing at 26°C for 24 hours, the crystals were
filtered off to give 0.15 g (79%) of Ca[(R)‐3)2(H‐DBTA)2]
with a de of 82% (Scheme 4, I). The diastereomeric
complex was purified further by stirring the diastereomer
in a mixture of 0.74 mL of ethanol and 0.74 mL of
ethyl acetate at 26°C for 24 hours to give 0.096 g (50%)
Ca[(R)‐3)2(H‐DBTA)2] with a de of 86% (Scheme 4, II).
The diastereomer was purified again as described above
to afford 0.027 g (14%) Ca[(R)‐3)2(H‐DBTA)2] with a de
of 87% (Table 2, entry 2). The (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] was recovered from
the diastereomer by treating the 2 mL dichloromethane
suspension of Ca[(R)‐3)2(H‐DBTA)2] with 2 mL of 10%
aqueous ammonia. The organic layer was washed with
0.5 mL of water, dried (Na2SO4), and concentrated to give
0.008 g (11%) of (R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3] with an ee of 87% (Scheme 4, III). All resolution
experiments of ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3) with Ca(H─DBTA)2 and Ca(H─DPTTA)2
[(−)‐6 and (−)‐7] were performed according to Repre-
sentative Procedure II. The conditions and the results
are shown in Table 2.
of the corresponding enantiomeric mixture of (R)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide [(R)‐3]. The crys-
tals that appeared on cooling the solution to 26°C were col-
lected by filtration after 15 minutes of crystallization.
Results: ee0: 23%: ee: 3% and yield: 57%; results: ee0: 41%:
ee: 6% and yield: 41%; results: ee0: 58%: ee: 61% and yield:
41%; results: ee0: 82%: ee: 97% and yield: 46% (Figure 1B).
(A)
(B)
2.5 | Preparation and recrystallization of
the enantiomeric mixtures of ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide (3)
Racemic and enantiopure (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [3 and (R)‐3] were mixed to pro-
vide a total amount of 0.10 g of the given enantiomeric
mixture. The exact enantiomeric composition of that given
enantiomeric mixture was determined by HPLC. In
0.15 mL of diethyl ether was dissolved 0.015 g (0.041 mmol)
FIGURE 1 A, The binary melting point and B, the ee0‐ee diagram
for ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3)

BAGI ET AL.
513
(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3] (ee > 99%)
2.6 | Comparative study on the
purification of non‐racemic mixtures of
ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3)
25
(Scheme 4, VII) [α]D25 = ½αꢀD +32.6 (c = 1.2 in CHCl3). Mp
94°C to 95°C.
2.6.1 | Repeated resolution according to
the half‐equivalent method
2.6.4 | Racemization of (R)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide
[(R)‐3] via chlorophosphonium salt (8)
To obtain 0.16 g (71%) (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] with an ee of 82%
(Scheme 4, IV), 0.45 g of Ca(R)‐3)2(H‐DBTA)2 (yield:
79%, de: 82%) prepared from 0.45 g (1.8 mmol) of racemic
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3), 0.69 g
(1.8 mmol) of DBTA·H2O, and 0.051 g (0.92 mmol) of CaO
as described in the Representative Procedure II was
decomposed by extraction with dichloromethane and
10% aqueous ammonia. The enantiomeric mixture so
obtained was resolved with 0.22 g (0.59 mmol) of
DBTA·H2O and 0.017 g (0.30 mmol) of CaO in a mixture
of 0.72 mL of ethyl acetate, 0.72 mL of ethanol, and
0.07 mL of water according to Representative Procedure
II. After the decomposition of Ca[(R)‐3)2(H‐DBTA)2]
diastereomer (Scheme 4, V), 0.084 g (38%) of (R)‐ethyl‐
(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3] was
obtained with an ee of 96%.
To the solution of 0.10 g (0.42 mmol) (R)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide [(R)‐3] (ee: 99%)
in 2 mL of dry dichloromethane was added dropwise
0.053 mL (0.62 mmol) of oxalyl chloride at 0°C over
10 minutes. The mixture was then allowed to warm to
26°C, and it was stirred under a nitrogen atmosphere for
2
hours. The solvent and the excess of the
reagent was removed under reduced pressure. The
chlorophosphonium salt so obtained was dissolved in
2 mL of dichloromethane. The mixture was cooled to
0°C, and 0.5 mL of water was added to the reaction
mixture over 10 minutes. The emulsion was stirred for
2 hours at 0°C. The 2 phases were separated, and the
aqueous layer was extracted with dichloromethane. The
organic layers were combined, and the resulting solution
was dried (Na2SO4); the solvent was evaporated, and the
crude product was passed through a silica gel column
(5% methanol in dichloromethane) to afford 0.082 g
(80%) of racemic ethyl‐(2‐methylphenyl)‐phenylphos-
phine oxide (3).
2.6.2 | Repeated resolution according to
the equivalent method
(R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3]
(0.16 g, ee: 82%) prepared as described in Section 2.6.1
was resolved with 0.25 g (0.65 mmol) of DBTA·H2O
and 0.018 g (0.33 mmol) of CaO in a mixture of 0.79 mL
of ethyl acetate, 0.79 mL of ethanol and 0.08 mL of
water according to Representative Procedure II. After
the decomposition of Ca[(R)‐3)2(H‐DBTA)2] diastereo-
2.7 | X‐ray measurements (general)
Intensity data were collected for the crystals of racemic
and
the
enantiopure
(R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [3 and (R)‐3] on an Xcalibur,
Sapphire3 diffractometer (graphite monochromator;
Mo‐Kα radiation, λ = 0.71069 Å) at 123(2)K. The data
collection was performed for the 1:1 molecular complex
of (R)‐3: spiro‐TADDOL on a D8 Venture diffractometer
(Mo‐Kα radiation, λ = 0.71069 Å) at 100 (2)K. All crys-
tals were mounted on glass fibers. Multiscan absorption
corrections were applied to the data in all cases. All ini-
tial structure models were obtained by direct methods
and subsequent difference syntheses.68 Models were then
refined by full‐matrix least squares procedures69 using
anisotropic displacement parameters for all non‐H
atoms. The standard treatment for hydrogen atomic
positions was calculations from assumed geometries,
unless indicated otherwise. Hydrogen atoms were
included in structure factor calculations, but they were
not refined, unless noted. The isotropic displacement
parameters of the hydrogen atoms were approximated
from the U(eq) value of the atom they were bonded.
mer, 0.093
g (42%) of (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] was obtained with an ee
of 96% (Scheme 4, VI). Mp of Ca[(R)‐3)2(H‐DBTA)2] (de:
96%): 169°C to 170°C.
2.6.3 | Recrystallization of the enantio-
meric mixture of (R)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide
[(R)‐3]
(R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3]
(0.16 g, ee: 82%) prepared as described in Section 2.6.1
was recrystallized from 0.40 mL of diethyl ether to
give 0.13 g (57 %) (R)‐ethyl‐(2‐methylphenyl)‐phenylphos-
phine oxide [(R)‐3] with an ee of 96%. Another recrystalli-
zation of this enantiomeric mixture from 0.40 mL of
diethyl ether afforded 0.10 g (47%) enantiopure (R)‐ethyl‐

514
BAGI ET AL.
Geometry calculations and validations were done.70 Part
of the drawings used in this work were made by the
perusal of DIAMOND71 and MERCURY.72
Final crystal structure model data are deposited with
the Cambridge Crystallographic Data Centre under
CCDC 1572195–1572197 and can be obtained free of
charge upon contacting.
3.2 | Resolution of ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) with TADDOL
derivatives (−)‐4 and (−)‐5
The resolution of ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3) was first attempted with TADDOL derivatives
(−)‐4 and (−)‐5. The phosphine oxide (3) and the corre-
sponding TADDOL derivative (−)‐4 or (−)‐5 were dis-
solved in boiling ethyl acetate. The precipitation of the
diastereomeric complexes having the composition of
(3)·(TADDOL) or (3)·(spiro‐TADDOL) was observed upon
the addition of hexane to the solution. The crystalline dia-
stereomers were separated by filtration after 3 hours, and
they were purified further by 2 recrystallizations. The
optically active ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3 or (S)‐3] could be liberated from the
corresponding diastereomer by column chromatography
(Scheme 2). The composition of the diastereomers was
3 | RESULTS AND DISCUSSION
3.1 | Preparation of racemic ethyl‐
(2‐methylphenyl)‐phenylphosphine
oxide (3)
The ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3)
was chosen as an air stable model compound having a
diaryl‐alkyl substitution pattern, as this organophospho-
rus compound (3) shows structural similarity to a
few P(III) ligands used in asymmetric syntheses.9,73
The title compound (3) was prepared by reacting
phenylphosphonic dichloride (1) with 2‐methylphenyl-
magnesium bromide. The phosphinic chloride intermediate
(2) was reacted further without isolation with ethyl-
magnesium bromide to afford ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) in a yield of 49% after the
work‐up and purification by column chromatography
(Scheme 1).
1
determined by H NMR, whereas the ee values of the
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) were
monitored by HPLC using a chiral stationary phase. The
results were collected in Table 1.
Despite our previous success,65 TADDOL [(−)‐4] was
not a suitable resolving agent for the enantioseparation
of the ethyl‐(2‐methylphenyl)‐phenylphosphine oxide
(3) as the overall efficiency for the resolution of phos-
phine oxide (3) remained low and the maximal ee was
only 3% (Table 1, entry 1). Although our initial
SCHEME 1 Synthesis of racemic
ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3)
SCHEME 2 General resolution procedure for racemic ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) using TADDOL derivatives (−)‐4
and (−)‐5

BAGI ET AL.
515
experiments
with
TADDOL
[(−)‐4]
remained
resolution of the title phosphine oxide (3). Both resolving
agents were prepared in situ by the reaction of (−)‐O,O′‐
dibenzoyl‐ or (−)‐O,O′‐di‐p‐toluoyl‐(2R,3R)‐tartaric acid
with CaO in a boiling mixture of ethanol and water.
To this solution was added ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) in ethyl acetate or in acetoni-
trile. The corresponding diastereomers precipitated from
the solution on cooling. After 24 hours, the crystals were
collected by filtration, and then they were purified further
by stirring the solid diastereomer in the corresponding
solvent mixture. The enantiomeric mixture of the phos-
phine oxide (3) could be liberated by adding aqueous
NH4OH to the diastereomeric mixture followed by extrac-
tion with dichloromethane (Scheme 3). The composition
of the diastereomers was analyzed by 1H NMR. The enan-
tiomeric purity was monitored by chiral HPLC. The
results are summarized in Table 2.
unsuccessful, the resolution of ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) was also attempted with
spiro‐TADDOL [(−)‐5], which is the structural analogue
of TADDOL [(−)‐4]. Applying half equivalent of spiro‐
TADDOL [(−)‐5] in the mixture of ethyl acetate and
hexane, the (R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3] could be obtained in an ee of 66% after
purification and decomposition of the corresponding dia-
stereomer (Table 1, entry 2).
Previously, we found that alcohols may also be suit-
able solvents for resolutions with TADDOL‐derivatives
(−)‐4 or (−)‐5, especially in the case of spiro‐TADDOL
[(−)‐5].65 Thus, the resolution of racemic phosphine oxide
(3) was also accomplished with 0.5 equivalents or 1 equiv-
alent of spiro‐TADDOL [(−)‐5] in isopropyl alcohol.
These experiments led to the best results (ee: 89%,
S = 0.38 and ee: 90%, S = 0.38) (Table 1, entries 5 and
6). Interestingly, the results obtained with the half equiv-
alent or the equivalent method showed parity.
The
resolution
of
ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) was first attempted with 0.25
or 0.5 equivalent of Ca(H‐DBTA)2 [(−)‐6] in a mixture of
ethyl acetate, ethanol, and water (Table 2, entries 1 and
2). In these experiments, coordination complexes with a
composition of Ca(3)2(H‐DBTA)2 were formed. In this
manner, resolutions were conducted according to the half
equivalent or the equivalent method, respectively (Table 2,
entries 1 or 2). The results obtained after the first crystal-
lization were promising, as the ee of (R)‐3 was 61%, and the
resolving capability was 0.37, when 0.25 equivalents of the
resolving agent [(−)‐6] were applied (Table 2, entry 1).
When both enantiomers were converted to the corre-
sponding diastereomer (ie, the resolution was conducted
according to the equivalent method), the enantiomeric
purity (ee) and the overall efficiency values increased to
82% and 0.64, respectively (Table 2, entry 2). After recrys-
tallizations, the purity of the diastereomers obtained in
these 2 experiments increased to 94% or 87%, respectively,
3.3 | Resolution of ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) with calcium
salts of the (−)‐O,O′‐dibenzoyl‐ or (−)‐O,O′‐
di‐P‐toluoyl‐(2R,3R)‐tartaric acids (−)‐4 or
(−)‐5
The TADDOL derivatives (−)‐4 or (−)‐5 were promising
resolving agents, as the (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] could be prepared with
an ee of 90%, but the overall efficiency of that resolution
experiment was moderate (S = 0.38) (Table 1, entry 4).
As the next step of this study, we turned our attention to
the application of the acidic calcium salts of the (−)‐O,O′‐
dibenzoyl‐ or (−)‐O,O′‐di‐p‐toluoyl‐(2R,3R)‐tartaric acid
[Ca(H‐DBTA)2 or Ca(H‐DPTTA)2; (−)‐6 or (−)‐7] for the
SCHEME 3 General resolution procedure for racemic ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) using Ca(H‐DBTA)2 or
Ca(H‐DPTTA)2 [(−)‐6 or (−)‐7]

516
BAGI ET AL.
but the resolving capability values decreased significantly
to 0.23 or 0.12, respectively (Table 2, entries 1 and 2),
which in fact is the consequence of the low yields (25%
or 14%, respectively) caused by the inefficient purification.
The enantioseparation of the title phosphine oxide (3)
was also attempted with Ca(H‐DBTA)2 [(−)‐6] in a mix-
ture of acetonitrile, ethanol, and water. However, this
change in the solvent mixture did not have a positive
impact on the results, as the enantiomeric excess values
showed parity, but the overall efficiency decreased (ee:
94%, S = 0.14) (Table 2, entry 3). The resolution of ethyl‐
(2‐methylphenyl)‐phenylphosphine oxide (3) was also
attempted with Ca(H‐DPTTA)2 [(−)‐7] using similar
molar ratios and solvents, as in the case of Ca(H‐DBTA)2
[(−)‐6]. Probably, as a consequence of the structural sim-
ilarity of the 2 resolving agents [(−)‐6 and (−)‐7], the
purity of the enantiomeric mixtures, as well as the overall
resolution efficiency could not be improved further with
Ca(H‐DPTTA)2 [(−)‐7], as the ee values fell in the range
of 82% to 88%, and the resolving capability values
were between 0.07 and 0.15 after purification (Table 2,
entries 4‐6).
compound having a eutectic composition (eeeu) of 52%
(Figure 1a).
To verify the accuracy of the calculated melting point
diagram, enantiomeric mixtures having the ee values of
approximately 20% to 40% to 60% to 80% were prepared,
and these samples were also subjected to DSC analysis
(See Supporting Information). The experimental melting
points of the enantiomeric mixtures showed good agree-
ment with the calculated values (Figure 1A).
Beside the ternary solubility diagrams, the ee0‐ee dia-
gram is a simple model to predict how a given enantio-
meric mixture behaves during crystallization.75 Thus, the
behavior of the enantiomeric mixtures of phosphine oxide
(3) was also elucidated by an ee0‐ee diagram. Separate
samples of enantiomeric mixtures prepared for DSC
measurements were recrystallized from diethyl ether. The
ee0‐ee diagram (Figure 1B), which was constructed by
plotting the enantiomeric excess after the recrystallization
(ee) against the initial enantiomeric purity (ee0) confirms
that the ethyl‐(2‐methylphenyl)‐phenylphosphine oxide
(3) is a racemic crystal‐forming compound, and the eutec-
tic composition determined by these recrystallization
experiments (eeeu: 56%) is also in good agreement with
the value suggested by the DSC measurements (eeeu: 52%).
Having all these results in hand, we wished to find the
most suitable method for the preparation of pure enantio-
mers. In the study of the resolving agents, the stirring of
the solid diastereomers in a given solvent mixture was
the purification method used. However, the purity of the
diasteremeric mixtures could only be improved by losing
a large amount of diastereomers during the purification,
which resulted in low yields, and consequently low effi-
ciency values (S). Besides the purification of diastereo-
mers used earlier (Table 2, entry 2), the repeated
resolution according to the half equivalent or the equiva-
lent method, or the recrystallization of the enantiomeric
mixtures was also considered as an option. The results
are summarized in Scheme 4. For comparable results, a
diastereomeric mixture obtained by the resolution with
Ca(H‐DBTA)2 according to the optimum conditions was
considered as the starting material (Table 2, entry 2).
Scheme 4 also illustrates the results shown in Table 2,
entry 2, and the corresponding enantiomeric mixture of
(R)‐3 could be obtained in a yield of 11% and with an ee
of 87% after decomposition of the diastereomer (Scheme 4,
III). In a separate experiment, the diastereomeric mixture
obtained after the first crystallization (Scheme 4, I) was
decomposed by extraction to obtain the enantiomeric
mixture of (R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3] in a yield of 71% and with an ee of 82%
(Scheme 4, IV). The purification of this enantiomeric
mixture was attempted by repeated resolution accor-
ding to the half equivalent or the equivalent method
3.4 | Purification of the enantiomeric
mixture of (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3]
Considering all of the results discussed in the previous
sections, we found that efficient enantiomeric separation
could be accomplished with Ca(H‐DBTA)2 [(−)‐6], but
pure enantiomers could not be obtained by the recrystalli-
zations of the corresponding diastereomers. Thus, other
purification strategies74,75 (eg, repeated resolution or
purification of enantiomeric mixtures) were evaluated
to prepare enantiopure (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3]. In our opinion, these
approaches are less discussed in the sphere of P‐stereo-
genic organophosphorus compounds.
First, the solid state behavior of the enantiomeric mix-
tures of ethyl‐(2‐methylphenyl)‐phenylphosphine oxide
(3) was investigated, as this information is essential to
design a proper purification method.75-77 Binary melting
point diagrams can be used to determine whether a given
enantiomeric mixture forms a conglomerate, racemic
crystals, or solid solution. Differential scanning calorime-
try measurements of the racemic compound and the pure
enantiomer were accomplished to determine the corre-
sponding melting point and heat of fusion values (See
Supporting Information), and the binary melting point
diagram was constructed using the simplified Schröder‐
van Laar and Prigogine‐Defay equations.76,77 The
results indicated that the ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide (3) is a racemic crystal‐forming

BAGI ET AL.
517
SCHEME 4 The purification procedure for the enantiomeric mixture of (R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3]
(Scheme 4, V or VI). The enantiomeric purity obtained
after the crystallization and decomposition of the diaste-
reomer was improved to 96% in both instances, but the
yields were moderate 38% or 42%, respectively (Scheme 4,
V or VI). Finally, the enantiomeric mixture was purified
without any chiral auxiliary, by a simple recrystallization
in diethyl ether. In this manner, enantiopure (R)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide [(R)‐3] (ee > 99%)
could be obtained in a yield of 47% (Scheme 4, VII). To
the best of our knowledge, despite the simplicity of this
method, only a few examples can be found in the litera-
ture, when the recrystallization of the enantiomeric mix-
tures was suitable for the preparation of enantiopure
P‐stereogenic organophosphorus compounds.21,44,78,79
racemization of phosphine oxide (3) was investigated as
the last step of this study. The racemic phosphine oxide
(3) so obtained from the undesired enantiomer could be
used as starting material in subsequent optimized resolu-
tion experiments, and hence, it may increase the overall
yield of a given resolution protocol.
In this study, the enantiopure (R)‐ethyl‐(2‐methylphenyl)‐
phenylphosphine oxide [(R)‐3] was considered as the
model compound.
First, the racemization of the enantiopure phosphine
oxide [(R)‐3] was attempted under thermal or microwave
conditions in a dimethylformamide solution or as a melt.
However, these racemization attempts were inefficient, as
enantiomeric mixtures of phosphine oxide (3) having an
ee of 95% could be obtained upon heating for extended
time (See Supporting Information). Recently, Gilheany
et al demonstrated that chlorophosphonium salts that
can be prepared from phosphines or phosphine oxides in
a single quantitative reaction step, racemize spontane-
ously, even at a low temperature.80 The racemization of
P‐stereogenic phosphine oxides via chlorophosphonium
3.5 | Racemization of optically active
ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide (3)
After a thorough investigation of the resolution of
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3), the

518
BAGI ET AL.
salt intermediates has already been used to epimerize the
undesired enantiomer obtained from a resolution process.81
Thus, the (R)‐ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3] was reacted with oxalyl‐chloride at 0°C
(Scheme 5), and then the chlorophosphonium salt (8)
so obtained was hydrolized to afford racemic ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide (3) in a yield
of 80%.
3.6 | Single crystal X‐ray analysis of
the racemic and enantiopure ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide
[3 and (R)‐3], as well as the (R)‐3‐spiro‐
TADDOL 1:1 complex
FIGURE
2
Asymmetric unit structure of the (R)‐3‐spiro‐
TADDOL 1:1 complex in the crystal. Non‐H atoms are shown in
50% probability of atomic displacement values
X‐ray quality crystals could be grown from racemic and
enantiopure
ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [3 and (R)‐3], as well as a diastereomeric complex
having the composition of (3)·(spiro‐TADDOL). The
crystallographic measurements enabled us to assign
the (R) absolute configuration to the (+)‐ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide [(+)‐3]. Crystal
structure models of both the racemic 3 (Figure S1) and
the enantiopure (R)‐3 (Figure S2) show practically identi-
cal shape and conformation of the 2 independent molecu-
lar structures (Figure S3).
The crystal structure of the (R)‐3‐spiro‐TADDOL com-
plex corroborates a 1:1 composition of the acyclic phos-
phine oxide with the spiro‐TADDOL host molecule
(Figure 2).
The comparison of the 2 (R)‐3 molecules observed in
the enantiopure phosphine oxide crystal and the one in
the 1:1 assembly with the spiro‐TADDOL [(−)‐5] host
shows only 1 marked difference (See Figure S4). Methyl
termini of the ethyl group is placed in an alternative
gauche position. Otherwise, the atomic positions more
or less coincide with the o‐methyl group exhibiting the
largest (approximately 0.6 Å) deviation. As this value is
below the experimental resolution, one may practically
take all atomic positions alike less the quoted methyl
termini.
The asymmetric unit drawing also shows the charac-
teristic intramolecular OH … O hydrogen bridge linking
the 2 ─OH functions. The only free ─OH acts as donor
to the O═P group, thus providing a basis for the host‐
guest assembly. This classical H‐bonding scheme is
complemented then by 4 short C—H … O contacts
(Table 3).
As an attempt to assess the relative interaction
strengths in the crystal, intermolecular potentials were
calculated,82,83 of which only the top 5 are shown
(Table 4). These identify that the principal interaction is
indeed between the ─OH acting as donor to the O═P
group of the target phosphine oxide (3). As these values
do not contain any error estimates, one can assume that
they are merely approximating the reality, and thus, serve
only as a guide. Nevertheless, as Figure S5 indicates, these
comply with the shortest interactions listed above,
and support their interpretation as the structure
directing ones. The calculated density of the complex
TABLE 3 Selected O—H … O and C—H … O short contact
dimensions (in Å and in °) with their standard deviations (σ) in
parentheses in the crystal structure of the (R)‐3‐spiro‐TADDOL 1:1
complex
Length, Length, Length, Angle,
Short Contacts
Å
Å
Å
°
O2—H2 … O3
0.91(3)
1.80(3)
2.706(2) 176(2)
2.633(2) 167(3)
2.964(2) 119(1)
2.749(2) 100(1)
2.808(2) 101(1)
2.914(2) 114(1)
O3—H3 … O1(x,y,z + 1) 0.91(3)
1.74(3)
2.41(1)
2.41(1)
2.48(1)
2.37(1)
C27—H27 … O4
C33—H33 … O3
C43—H43 … O2
C45—H45 … O5
0.92(1)
0.96(2)
0.95(1)
0.99(2)
SCHEME 5 The racemization of (R)‐
ethyl‐(2‐methylphenyl)‐phenylphosphine
oxide [(R)‐3] via chlorphosphonium
salt (8)

BAGI ET AL.
519
TABLE 4 Calculated intermolecular potentials (packing energy)
rounded to integer numbers
of P‐stereogenic phosphine oxides. Moreover, we also
demonstrated that the enantiomeric mixtures of P‐chiral
phosphine oxides can be purified efficiently by a simple
recrystallization, which method is rarely used in the
sphere of P‐chiral compounds. This approach requires
the knowledge of the solid state behavior of the particular
chiral compound. However, this rather simple purifica-
tion can serve as a supplementary method to purify
non‐racemic mixtures of phosphine oxides obtained by
inefficient resolution methods or asymmetric syntheses.
Mol 1
Mol 2
Distance
Energy, kJ/mol
0
1
7.13249
−56
2
0
0
0
3
4
5
6
7.13249
8.621
−56
−39
−39
−36
8.621
8.91578
(1.24 Mg/m3) gets closer to that of the racemic crystal, yet
another indication of the importance of attractive
interactions.
ACKNOWLEDGMENTS
This work was supported by the National Research,
Development and Innovation Office—NKFIH (Grant
No. OTKA PD 116096). PB thanks the financial support
of the ÚNKP‐17‐4‐I‐BME‐102 New National Excellence
Program of the Ministry of Human Capacities.
4 | CONCLUSIONS
In summary, a detailed study was conducted to investi-
gate all main aspects of the resolution procedures of
acyclic diaryl‐alkylphosphine oxides. The ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide (3) was considered
as the model compound, and the resolution of phosphine
oxide (3) was elaborated with TADDOL derivatives [(−)‐4
or (−)‐5] and the acidic calcium salt of the (−)‐O,O′‐
dibenzoyl‐ or (−)‐O,O′‐di‐p‐toluoyl‐(2R,3R)‐tartaric acid
[(−)‐6 or (−)‐7]. Both resolving agent classes were
applicable for the enantiomeric separation of phosphine
oxide (3), and the best results could be obtained using
Ca(H‐DBTA)2 (−)‐6. The purification strategies of the
enantiomerically enriched phosphine oxide (3) were
investigated in detail. We found that enantiopure (R)‐
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide [(R)‐3]
could be prepared by resolution with Ca(H‐DBTA)2 [(−)‐6]
followed by recrystallization of the partially enantioen-
riched phosphine oxide (3). An efficient racemization of
ethyl‐(2‐methylphenyl)‐phenylphosphine oxide (3) enan-
tiomers via the formation of chlorophosphonium salt (8)
was also elaborated. In this manner, the undesired enan-
tiomer of the phosphine oxide (3) may be converted into
racemic 3 that can be used in subsequent resolution
experiments as a starting material. Single crystal X‐ray
analysis of the 1:1 diastereomeric coordination complex
(R)‐3 : spiro‐TADDOL point to the fact that the principal
molecular recognition events are affected by the ─OH
group acting as donor to the O═P group of the target
phosphine oxide (3).
ORCID
Péter Bagi http://orcid.org/0000-0002-9043-6435
REFERENCES
1. Börner A (Ed). Phosphorus Ligands in Asymmetric Catalysis.
Weinheim: Wiley‐VCH; 2008.
2. Ojima I (Ed). Catalytic Asymmetric Synthesis. Weinheim: Wiley‐
VCH; 2010.
3. Pietrusiewicz KM, Zablocka M. Preparation of Scalemic P‐Chiral
Phosphines and Their Derivatives. Chem Rev. 1994;94:1375‐1411.
4. Grabulosa A. P‐Stereogenic Ligands in Enantioselective Catalysis.
Cambridge: The Royal Society of Chemistry; 2010.
5. Herault D, Nguyen DH, Nuel D, Buono G. Reduction of second-
ary and tertiary phosphine oxides to phosphines. Chem Soc Rev.
2015;44:2508‐2528.
6. Chaux F, Frynas S, Laureano H, et al. Enantiodivergent synthe-
sis of P‐chirogenic phosphines. C R Chim. 2010;13:1213‐1226.
7. Grabulosa A, Granell J, Muller G. Preparation of optically pure
P‐stereogenic trivalent phosphorus compounds. Coord Chem
Rev. 2007;251:25‐90.
8. Kolodiazhnyi OI, Kukhar VP, Kolodiazhna AO. Asymmetric
catalysis as a method for the synthesis of chiral organophospho-
rus compounds. Tetrahedron: Asymmetry. 2014;25:865‐922.
9. Dutartre M, Bayardon J, Jugé S. Applications and stereoselective
syntheses of P‐chirogenic phosphorus compounds. Chem Soc
Rev. 2016;45:5771‐5794.
This study is the first example for the application of
TADDOL derivatives [(−)‐4 or (−)‐5] and the calcium
salts of tartaric acid derivatives [(−)‐6 or (−)‐7] for the res-
olution of an acyclic phosphine oxide. Our experience
indicates that these enantiomeric separation methods
may have a more general value during classical resolution
10. Korpiun O, Lewis RA, Chickos J, Mislow K. Synthesis and
absolute configuration of optically active phosphine oxides and
phosphinates. J Am Chem Soc. 1968;90:4842‐4846.
11. Gatineau D, Giordano L, Buono G. Bulky, optically active P‐stereo-
genic phosphine–boranes from pure H‐menthylphosphinates.
J Am Chem Soc. 2011;133:10728‐10731.

520
BAGI ET AL.
12. Berger O, Montchamp JL. A general strategy for the synthesis of
P‐stereogenic compounds. Angew Chem Int Ed. 2013;52:11377‐
11380.
mandatory for reports involving scalemates? Tetrahedron: Asym-
metry. 2017;28:1430‐1434.
27. Cox GB (Ed). Preparative Enantioselective Chromatography.
Oxford: Blackwell Publishing; 2005.
13. Wang WM, Liu LJ, Zhao CQ, Han LB. Diastereoselective hydro-
lysis of asymmetric P‐Cl species and synthesis of optically pure
28. Jiang X, Minnaard AJ, Hessen B, et al. Application of
monodentate secondary phosphine oxides, a new class of chiral
ligands, in Ir(I)‐catalyzed asymmetric imine hydrogenation.
Org Lett. 2003;5:1503‐1506.
(RP)‐(−)‐menthyl H‐phenylphosphinate. Eur
2015;2015:2342‐2345.
J Org Chem.
14. Berger O, Montchamp JL. General synthesis of P‐stereogenic
compounds: the menthyl phosphinate approach. Org Biomol
Chem. 2016;1:7552‐7562.
29. Jiang X, van den Berg M, Minnaard AJ, Feringa BL, de Vries JG.
The application of monodentate secondary phosphine oxide
ligands in rhodium‐ and iridium‐catalyzed asymmetric hydroge-
nation. Tetrahedron: Asymmetry. 2004;15:2223‐2229.
15. Jugé S, Genet JP. Asymmetric synthesis of phosphinates, phos-
phine oxides and phosphines by Michaelis Arbuzov
rearrangement of chiral oxazaphospholidine. Tetrahedron Lett.
1989;30:2783‐2786.
30. Imamoto T, Crépy KV, Katagiri K. Optically active 1,1′‐di‐tert‐
butyl‐2,2′‐dibenzophosphetenyl:
a
highly
strained
P‐
stereogenic diphosphine ligand. Tetrahedron: Asymmetry.
2004;15:2213‐2218.
16. Jugé S, Stephan M, Laffitte JA, Genet JP. Efficient asymmetric
synthesis of optically pure tertiary mono and diphosphine
ligands. Tetrahedron Lett. 1990;31:6357‐6360.
31. Bigeault J, Giordano L, Buono G. [2+1] Cycloadditions of termi-
nal alkynes to norbornene derivatives catalyzed by palladium
complexes with phosphinous acid ligands. Angew Chem Int Ed.
2005;44:4753‐4757.
17. Carey JV, Barker MD, Brown JM, Russell MJH. Preparation of
enantiomerically pure phosphine oxides by nucleophilic dis-
placement chemistry using oxazaphospholidines. J Chem Soc
Perkin Trans 1. 1993;7:831‐839.
32. Widhalm M, Brecker L, Mereiter K. Rigid P‐chiral mono and
diphosphines. Configurative stability and P‐inversion barrier.
Tetrahedron: Asymmetry. 2006;17:1355‐1369.
18. Oliana M, King F, Horton PN, Hursthouse MB, Hii KK. Practical
synthesis of chiral vinylphosphine oxides by direct nucleophilic
substitution. Stereodivergent synthesis of aminophosphine
ligands. J Org Chem. 2006;71:2472‐2479.
33. Gatineau D, Nguyen DH, Hérault D, et al. H‐
adamantylphosphinates as universal precursors of P‐stereogenic
compounds. J Org Chem. 2015;80:4132‐4141.
19. Copey L, Jean‐Gérard L, Andrioletti B, Framery E. Synthesis of
P‐stereogenic secondary phosphine oxides using α‐d‐glucos-
amine as a chiral precursor. Tetrahedron Lett. 2016;57:543‐545.
34. Leek H, Thunberg L, Jonson AC, Öhlén K, Klarqvist M. Strategy
for large‐scale isolation of enantiomers in drug discovery. Drug
Discov Today. 2017;22:133‐139.
20. Han ZS, Goyal N, Herbage MA, et al. Efficient asymmetric syn-
thesis of P‐chiral phosphine oxides via properly designed and
activated benzoxazaphosphinine‐2‐oxide agents. J Am Chem
Soc. 2013;135:2474‐2477.
35. Tang W, Qu B, Capacci AG, et al. Novel, tunable, and efficient
chiral bisdihydrobenzooxaphosphole ligands for asymmetric
hydrogenation. Org Lett. 2010;12:176‐179.
21. Han ZS, Wu H, Xu Y, et al. General and stereoselective method
for the synthesis of sterically congested and structurally diverse
P‐stereogenic secondary phosphine oxides. Org Lett.
2017;19:1796‐1799.
36. Andersen NG, Ramsden PD, Che D, Parvez M, Keay BA. A novel
resolution procedure for the preparation of P‐stereogenic phos-
phine oxides. Org Lett. 1999;1:2009‐2011.
37. Dziuba K, Flis A, Szmigielska A, Pietrusiewicz KM. Efficient oxi-
22. Soloshonok V, Roussel C, Kitagawa O, Sorochinsky A. Self‐dis-
proportionation of enantiomers via achiral chromatography: a
warning and an extra dimension in optical purifications. Chem
Soc Rev. 2012;41:4180‐4188.
dative resolution of
asymmetric synthesis of
biphenyl diphosphine dioxide. Tetrahedron: Asymmetry.
2010;21:1401‐1405.
a
P‐stereogenic triarylphosphine and
P‐stereogenic atropoisomeric
a
23. Soloshonok V, Klika K. Terminology related to the phenomenon
‘self‐disproportionation of enantiomers’ (SDE). Helv Chim Acta.
2014;97:1583‐1589.
38. Toda F, Mori K, Zafra S, Goldberg I. Optical resolution of
phosphinates and phosphine oxides by complex formation with
optically active 2,2′‐dihydroxy‐1,1′‐binaphthyl and crystallo-
graphic study of two diastereomeric complexes with (CH3)
(C6H5)(OCH3)PO. J Org Chem. 1988;53:308‐312.
24. Wzorek A, Klika KD, Drabowicz J, Sato A, Aceña JL,
Soloshonok V. The self‐disproportionation of the enantiomers
(SDE) of methyl n‐pentyl sulfoxide via achiral, gravity‐driven
column chromatography: a case study. Org Biomol Chem.
2014;12:4738‐4746.
39. Drabowicz J, Łyżwa P, Omelańczuk J, Pietrusiewicz KM,
Mikołajczyk M. New procedures for the resolution of chiral
tert‐butylphenylphosphine oxide and some of its reactions. Tet-
rahedron: Asymmetry. 1999;10:2757‐2763.
25. Suzuki Y, Han J, Kitagawa O, Aceña JL, Klika KD, Soloshonok
VA. comprehensive examination of the self‐disproportionation
of enantiomers (SDE) of chiral amides via achiral, laboratory‐
routine, gravity‐driven column chromatography. RSC Adv.
2015;5:2988‐2993.
40. Bergin E, O'Connor CT, Robinson SB, McGarrigle EM,
O'Mahony CP, Gilheany DG. Synthesis of P‐stereogenic phos-
phorus compounds. Asymmetric oxidation of phosphines under
Appel conditions. J Am Chem Soc. 2007;129:9566‐9567.
26. Soloshonok V, Wzorek A, Klika KD. A question of policy: should
41. Rajendran KV, Gilheany DG. Identification of a key intermedi-
ate in the asymmetric Appel process: one pot stereoselective
tests for the self‐disproportionation of enantiomers (SDE) be

BAGI ET AL.
521
synthesis of P‐stereogenic phosphines and phosphine boranes
from racemic phosphine oxides. Chem Commun.
2012;48:10040‐10042.
complex formation with TADDOL derivatives. Tetrahedron:
Asymmetry. 2006;17:2599‐2602.
56. Novák T, Ujj V, Schindler J, et al. Resolution of 1‐substituted‐3‐
methyl‐3‐phospholene 1‐oxides by molecular complex formation
42. Nikitin K, Rajendran KV, Müller‐Bunz H, Gilheany DG. Turn-
ing regioselectivity into stereoselectivity: efficient dual
resolution of P‐stereogenic phosphine oxides through bifurca-
tion of the reaction pathway of a common intermediate. Angew
Chem Int Ed. 2014;53:1906‐1909.
with
TADDOL
derivatives.
Tetrahedron:
Asymmetry.
2007;18:2965‐2972.
57. Ujj V, Schindler J, Novák T, Czugler M, Fogassy E, Keglevich G.
Coordinative resolution of 1‐phenyl‐ and 1‐naphthyl‐3‐methyl‐3‐
phospholene 1‐oxides with calcium hydrogen O,O′‐dibenzoyl‐
(2R,3R)‐tartrate or calcium hydrogen O,O′‐di‐p‐toluyl‐(2R,3R)‐
tartrate. Tetrahedron: Asymmetry. 2008;19:1973‐1977.
43. Rajendran KV, Nikitin KV, Gilheany DG. Hammond postulate
mirroring enables enantiomeric enrichment of phosphorus com-
pounds via two thermodynamically interconnected sequential
stereoselective processes. J Am Chem Soc. 2015;137:9375‐9381.
58. Ujj V, Bagi P, Schindler J, Madarász J, Fogassy E, Keglevich G. A
practical and efficient method for the resolution of 3‐
phospholene 1‐oxides via coordination complex formation. Chi-
rality. 2010;22:699‐705.
44. Kortmann FA, Chang MC, Otten E, Couzijn EPA, Lutz M,
Minnaard AJ. Consecutive dynamic resolutions of phosphine
oxides. Chem Sci. 2014;5:1322‐1327.
45. Miura T, Imamoto T. Enantiomerically pure 1,2‐
bis(isopropylmethylphosphino)benzene and its use in highly
enantioselective Rh‐catalyzed asymmetric hydrogenation. Tetra-
hedron Lett. 1999;40:4833‐4836.
59. Bagi P, Laki A, Keglevich G. Preparation of optically active six‐
membered p‐heterocycles: a 3‐phosphabicyclo[3.1.0] hexane 3‐
oxide,
a 1,2‐dihydrophosphinine 1‐oxide, and a 1,2,3,6‐
tetrahydrophosphinine 1‐oxide. Heteroat Chem. 2013;24:179‐186.
46. Omelańczuk J, Karaçar A, Freytag M, et al. Chiral 1,8‐bis(tert‐
butylphenylphosphino)naphthalene oxides and sulfides: resolu-
tion and structures. Inorg Chim Acta. 2003;350:583‐591.
60. Bagi P, Kállay M, Hessz D, et al. Resolution of 1‐n‐propoxy‐3‐
methyl‐3‐phospholene 1‐oxide by diastereomeric complex for-
mation using TADDOL derivatives and calcium salts of O,O′‐
dibenzoyl‐(2R,3R)‐ or O,O′‐di‐p‐toluoyl‐(2R,3R)‐tartaric acid.
Tetrahedron: Asymmetry. 2014;25:318‐326.
47. Liu D, Gao W, Wang C, Zhang X. Practical synthesis of
enantiopure γ‐amino alcohols by rhodium‐catalyzed asymmetric
hydrogenation of β‐secondary‐amino ketones. Angew Chem Int
Ed. 2005;44:1687‐1689.
61. Bagi P, Fekete A, Kállay M, et al. Resolution of 1‐n‐butyl‐3‐
methyl‐3‐phospholene 1‐oxide with TADDOL derivatives and
calcium salts of O,O′‐dibenzoyl‐(2R,3R)‐ or O,O′‐di‐p‐toluoyl‐
(2R,3R)‐tartaric acid. Chirality. 2014;26:174‐182.
48. Takaya H, Mashima K, Koyano K, et al. Practical synthesis of
(R)‐ or (S)‐2,2′‐bis(diarylphosphino)‐1,1′‐binaphthyls (BINAPs).
J Org Chem. 1986;51:629‐635.
62. Bagi P, Fekete A, Kállay M, et al. A case study on the resolution
of the 1‐i‐butyl‐3‐methyl‐3‐phospholene 1‐oxide via diastereo-
meric complex formation using TADDOL derivatives and via
diastereomeric coordination complexes formed from the calcium
salts of O, O′‐diaroyl‐(2R,3R)‐tartarates. Heteroat Chem.
2015;26:79‐90.
49. Minami T, Okada Y, Nomura R, Hirota S, Nagahara Y, Fuku-
yama K. Synthesis and resolution of a new type of chiral
bisphosphine
ligand,
trans‐bis‐1,2‐(diphenylphosphino)
cyclobutane, and asymmetric hydrogenation using its rhodium
complex. Chem Lett. 1986;4:613‐616.
63. Bagi P, Juhász K, Timári I, et al. A study on the optical resolu-
tion of 1‐isopropyl‐3‐methyl‐3‐phospholene 1‐oxide and its use
in the synthesis of borane and platinum complexes. J Organomet
Chem. 2015;797:140‐152.
50. Benincori T, Gladiali S, Rizzo S, Sannicolo FA. New modular
class of easily accessible, inexpensive, and efficient chiral
diphosphine ligands for homogeneous stereoselective catalysis.
J Org Chem. 2001;66:5940‐5942.
64. Bagi P, Karaghiosoff K, Czugler M, et al. Synthesis, characteriza-
tion, and application of platinum(II) complexes incorporating
racemic and optically active 4‐chloro‐5‐methyl‐1‐phenyl‐1,2,3,6‐
tetrahydrophosphinine ligand. Heteroat Chem. 2016;27:91‐101.
51. Tang W, Chi Y, Zhang X. An ortho‐substituted BIPHEP ligand
and its applications in Rh‐catalyzed hydrogenation of cyclic
enamides. Org Lett. 2002;4:1695‐1698.
52. de Paule SD, Jeulin S, Ratovelomanana‐Vidal V, Genêt JP,
Champion N, Dellis P. Synthesis and molecular modeling stud-
ies of SYNPHOS, a new, efficient diphosphane ligand for
ruthenium‐catalyzed asymmetric hydrogenation. Eur J Org
Chem. 2003;2003:1931‐1941.
65. Bagi P, Ujj V, Czugler M, Fogassy E, Keglevich G. Resolution of
P‐stereogenic P‐heterocycles via the formation of diastereomeric
molecular and coordination complexes (a review). Dalton Trans.
2016;45:1823‐1842.
66. Beck AK, Bastani B, Plattner DA, et al. Large‐scale preparation
of α, α, α′, α′‐tetraaryl‐1,3‐dioxolane‐4,5‐dimethanols
(TADDOL)—versatile auxiliaries for EPC synthesis and its
solid‐state structure. Chimia. 1991;45:238‐244.
53. Rankic DA, Parvez M, Keay BA. 3,3′‐Substituted BINAP deriva-
tives containing C‐bound substituents: applications in
asymmetric hydrogenation reactions. Tetrahedron: Asymmetry.
2012;23:754‐763.
67. Sheldon RA. Chirotechnology: Industrial Synthesis of Optically
Active Compounds. New York: Marcel Dekker; 1993.
54. Holt J, Maj AM, Schudde EP, et al. On the resolution of second-
ary phosphine oxides via diastereomeric complex formation: the
case
of
tert‐butylphenylphosphine
oxide.
Synthesis.
68. Sheldrick GM. A short history of SHELX. Acta Cryst. A.
2008;64:112‐122.
2009;2009:2061‐2065.
55. Novák T, Schindler J, Ujj V, Czugler M, Fogassy E, Keglevich G.
69. Sheldrick GM. Crystal structure refinement with SHELXL. Acta
Cryst C. 2015;71:3‐8.
Resolution of 3‐methyl‐3‐phospholene 1‐oxides by molecular

522
BAGI ET AL.
70. Spek AL. Structure validation in chemical crystallography. Acta
Cryst D. 2009;65:148‐155.
80. Jennings EV, Nikitin K, Ortin Y, Gilheany DG. Degenerate
nucleophilic substitution in phosphonium salts. J Am Chem
Soc. 2014;136:16217‐16226.
71. Pennington WT. DIAMOND—visual crystal structure informa-
tion system. J Appl Cryst. 1999;32:1028‐1029.
81. Li G, Wang X, Zhang Y, et al. Synthesis of P‐chiral
dihydrobenzooxaphosphole core for BI ligands in asymmetric
transformations. J Org Chem. 2017;82:5456‐5460.
72. Macrae CF, Bruno IJ, Chisholm JA, et al. 2.0—new features for
the visualization and investigation of crystal structures. J Appl
Cryst. 2008;41:466‐470.
82. Gavezzotti A. Are crystal structures predictable? Acc Chem Res.
1994;27:309‐314.
73. Lagasse F, Kagan HB. Chiral monophosphines as ligands for
asymmetric organometallic catalysis. Chem Pharm Bull.
2000;48:315‐324.
83. Gavezzotti A, Filippini G. Geometry of the intermolecular X‐H.
cntdot..cntdot..cntdot.Y (X, Y = N, O) hydrogen bond and the
calibration of empirical hydrogen‐bond potentials. J Phys Chem.
1994;98:4831‐4837.
74. Faigl F, Fogassy E, Nógrádi M, Pálovics E, Schindler J. Strategies
in optical resolution: a practical guide. Tetrahedron: Asymmetry.
2008;19:519‐536.
75. Faigl F, Fogassy E, Nógrádi M, Pálovics E, Schindler J. Separa-
tion of non‐racemic mixtures of enantiomers: an essential part
of optical resolution. Org Biomol Chem. 2010;8:947‐959.
SUPPORTING INFORMATION
Additional Supporting Information may be found online
in the supporting information tab for this article.
76. Lorenz H, Seidel‐Morgenstern A. Binary and ternary phase dia-
grams of two enantiomers in solvent systems. Thermochim
Acta. 2002;382:129‐142.
77. Lorenz H, Seidel‐Morgenstern A. Processes to separate enantio-
mers. Angew Chem Int Ed. 2014;53:1218‐1250.
How to cite this article: Bagi P, Varga B, Szilágyi
A, et al. The resolution of acyclic P‐stereogenic
phosphine oxides via the formation of
diastereomeric complexes: A case study on ethyl‐(2‐
methylphenyl)‐phenylphosphine oxide. Chirality.
2018;30:509–522. https://doi.org/10.1002/
chir.22816
78. Serreqi AN, Kazlauskas RJ. Kinetic resolution of phosphines and
phosphine oxides with phosphorus stereocenters by hydrolases. J
Org Chem. 1994;59:7609‐7615.
79. Leyris A, Nuel D, Giordano L, Achard M, Buono G.
Enantioselective synthesis of both enantiomers of tert‐
butylphenylphosphine oxide from (S)‐prolinol. Tetrahedron Lett.
2005;46:8677‐8680.